Fiber-optic fallout

Billions were wasted in frenzy to build
networks, 90% of which lie dormant

By Jeff Smith, Rocky Mountain News Staff Writer

More than 18 months into the telecom
downturn, experts are still divided over the
depth of the bandwidth glut.

But there’s little disagreement about the
money wasted building long-distance
fiber-optic networks to carry high-speed
communications traffic.

A Boston consulting firm, Adventis Corp., recently calculated that about $70
billion was squandered over the past five years in the "crazy" frenzy to build
Internet backbones across the country.

"It was very short-sighted" and much of the fiber may never be used, said
Blake Kirby, vice president of Adventis Corp. About $139 billion was spent
wisely, according to the company.

The Oklahoma communications consulting firm TeleChoice has a slightly
different take.

"The bottom line isn’t that there’s too much fiber, the problem is that there are
too many companies that put that fiber into the ground," said Russ McGuire,
TeleChoice’s chief strategy officer. "It was a fundamental capital market
mistake. I don’t know of any other time when capital markets have funded
more than three (players) in any given utility market. It’s just dumb."

On some routes as many as 10 to 15 companies have built backbones.
"Denver to Chicago has 800 fibers," McGuire said. "That’s about the right
number (for long-term future needs). The problem is that probably eight
companies (including Denver-based Qwest Communications and
Broomfield-based Level 3 Communications) built those 800 fibers. There
should have been only two or three."

The common thread between these two views is this: Much of the money
wasted wasn’t in fiber, but in construction crews, engineers, executives,
buildings and everything else needed to fund such telecom upstarts as
Qwest and Level 3.

"Each of these companies spent $10 billion to $20 billion to build out their
national networks," Kirby said. "It doesn’t matter whether you light the fiber
now or later, you aren’t going to generate the revenue to pay off the debt."

The 1996 Telecommunications Act, which was supposed to usher in a new
era of competition, instead spurred a disastrous network overbuild.

Adventis’ chief executive Mark Bruneau concluded back in 1999 that there
would be a capacity glut. "But at the time, we were viewed as heretics," Kirby
said.

And amid the dot-com boom, more capital poured in.

Historically, Kirby said, about $3 has been spent on the local
telecommunications infrastructure for every dollar spent on the long-distance
backbone. But that formula became inverted in recent years, with $3 being
spent on the backbone for every dollar spent in local access.

Because of that inversion, the so-called "last mile" bottleneck became even
more pronounced. Most U.S. consumers still access the Internet through
dial-up modems rather than high-speed broadband connections and that
isn’t expected to change anytime soon.

"Level 3’s original business plan was to make $2.9 billion in revenues this
year, and they’re at half of that," Kirby said. "Unit volumes may be increasing,
but the revenue being spent on bandwidth right now is about the same (as)
three or four years ago. The reason is that it’s getting cheaper and cheaper to
deploy (bandwidth) every year. There’s newer fiber, newer electronics; it just
gets cheaper."

Adventis says that less than 10 percent of the fiber laid in the ground has
been "lit" with the amplifiers, routers and other communications devices to
make it operational.

McGuire, who has looked at 22 long-distance routes connecting the country’s
top 12 cities, agrees there’s plenty of fiber in the ground now. But he said
that’s not the point.

. "It takes a long time to put the fiber in the ground," so companies want to
plan 20 years in advance. "Whether it’s the right amount, too much or too
little, is kind of hard to know because I don’t want somebody to be digging in
my back yard again for a very long time." By some scenarios, McGuire said,
only 30 percent of the fiber now in the ground will be lit by 2005; other
scenarios "have it all lit by 2005."

In gauging whether there’s a bandwidth glut, it’s more important to look at the
"lit" fiber, experts say. Decisions are much more deliberate, because it costs
at least eight times as much to light fiber as lay fiber.

McGuire said his analysis is that 18 of the 22 major long-distance routes are
operating near full capacity when looking at "lit" fiber and only four routes
have excess capacity.

In a report late last month, the Washington, D.C., consulting firm
TeleGeography Inc. didn’t agree, but it did say that the dramatic price
collapse caused by the bandwidth glut may be nearing an end.

TeleGeography said the supply of the city-to-city bandwidth still "far exceeds
actual needs." But with prices already near or below cost, "it seems unlikely
that the capacity oversupply will depress prices any further." Carriers
basically can’t cut the prices any more without risking bankruptcy, according
to TeleGeography analyst Stephan Beckert.

For example, in 2000, an OC-3 circuit between New York and Los Angeles —
capable of simultaneously handling 2,400 conversations — had a pricetag of
$1.8 million a year. In the first quarter of 2002, the same annual lease cost
less than $150,000.

Whether bandwidth glut or not, what is the future for an industry that has
wasted tens of billions of dollars and can’t repay its debt?

Experts agree the most likely scenario is continued consolidation until there
are only a few players remaining in the long-haul market.

"A company probably needs at least a 10 to 20 percent market share to have
enough revenues and economies of scale" to pay off its initial investment,
McGuire said. And, as he noted, currently "seventy percent of the market is
locked up by Sprint, WorldCom, AT&T."

Level 3 can take comfort in the recent bankruptcies of rivals such as
360Networks, Global Crossing and Williams Communications. Yet
debt-laden Level 3 could be in trouble if some of these companies are able
to emerge from bankruptcy reorganization debt-free.

"Level 3 has the right business model to eventually make money," Kirby said.
"They have a national conduit infrastructure (and) they can upgrade it at a
pretty cheap price. In Level 3’s defense, they certainly have a war chest (more
than $1 billion in cash)."

The question is whether there will be enough demand, soon enough.
"Fundamentally, the glut has to wear out of the industry," Kirby said. "When
Nortel and Lucent start selling a lot more (telecom) gear and Cisco more
routers and AT&T and WorldCom start growing revenues, then the outlook for
Level 3 looks great."

Fiber-optic rival Qwest, meanwhile, he said, "did the right things at the right
time," acquiring U S West, a regional phone company, in mid-2000, before
the telecom downturn. "Fundamentally, that provides a lot of cash flow to get
through the storm," Kirby said, although Qwest also is facing the problem of
a high debt load and a federal investigation of its accounting practices.

McGuire said he believed the best outcome would be if Level 3, Williams,
Global Crossing and Broadwing merged. "The combination of those four
companies would be extremely powerful."

http://www.rockymountainnews.com/drmn/technology/article/0,1299,DRMN_49_1126487,00.html